purab manglani

I’m a Big Data Analytics master’s student building end-to-end solutions that turn messy data into reliable predictions and usable products. I combine strong programming foundations with hands-on ML and practical web/cloud experience, always aiming for measurable performance and maintainable delivery.

In my projects, I engineered features and evaluated models rigorously—like achieving R² = 0.88 with Random Forest for NYC 311 service request resolution time prediction using GridSearchCV and cyclical time encoding. I also built an end-to-end laptop price prediction pipeline (R² = 0.78) with log transforms, outlier handling, and defenses against leakage and overfitting, and I developed a deep-learning breast cancer classifier using TensorFlow/Keras with early stopping and dropout.

During internships, I shipped web and mobile applications using React.js, Next.js, and React Native, and I managed deployment pipelines on AWS to support scalability and uptime. I also worked with PHP frameworks (Laravel/CodeIgniter) and React to deliver client-ready sites—my code quality and delivery led to a formal extension opportunity.